Mask-Predict: Parallel Decoding of Conditional Masked Language Models

Marjan Ghazvininejad, Omer Levy, Yinhan Liu, Luke Zettlemoyer


Abstract
Most machine translation systems generate text autoregressively from left to right. We, instead, use a masked language modeling objective to train a model to predict any subset of the target words, conditioned on both the input text and a partially masked target translation. This approach allows for efficient iterative decoding, where we first predict all of the target words non-autoregressively, and then repeatedly mask out and regenerate the subset of words that the model is least confident about. By applying this strategy for a constant number of iterations, our model improves state-of-the-art performance levels for non-autoregressive and parallel decoding translation models by over 4 BLEU on average. It is also able to reach within about 1 BLEU point of a typical left-to-right transformer model, while decoding significantly faster.
Anthology ID:
D19-1633
Volume:
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)
Month:
November
Year:
2019
Address:
Hong Kong, China
Editors:
Kentaro Inui, Jing Jiang, Vincent Ng, Xiaojun Wan
Venues:
EMNLP | IJCNLP
SIG:
SIGDAT
Publisher:
Association for Computational Linguistics
Note:
Pages:
6112–6121
Language:
URL:
https://aclanthology.org/D19-1633
DOI:
10.18653/v1/D19-1633
Bibkey:
Cite (ACL):
Marjan Ghazvininejad, Omer Levy, Yinhan Liu, and Luke Zettlemoyer. 2019. Mask-Predict: Parallel Decoding of Conditional Masked Language Models. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 6112–6121, Hong Kong, China. Association for Computational Linguistics.
Cite (Informal):
Mask-Predict: Parallel Decoding of Conditional Masked Language Models (Ghazvininejad et al., EMNLP-IJCNLP 2019)
Copy Citation:
PDF:
https://aclanthology.org/D19-1633.pdf
Code
 facebookresearch/Mask-Predict +  additional community code